Morality refers to the set of values that people use to determine appropriate behavior, that is, what is right versus what is wrong. Determining which behavior is morally appropriate, or ”right,” is essentially a cognitive decision-making process called moral judgment.
Moral judgment is but one component of the process leading to the actual performance of morally appropriate behavior (Rest 1986). However, research on moral development over the past forty-five years has focused primarily on the development of moral judgment. This is due in large part to the influence of psychologists Lawrence Kohlberg (1969, 1971, 1976) and Jean Piaget ([1932] 1948).
Both Piaget and Kohlberg maintained that moral behavior largely depends upon how one perceives the social world and oneself in relation to it. Furthermore, they viewed moral decision making as a rational process and thus linked the development of moral judgment to the development of rational cognition. In this way, moral development is seen largely as changes in one’s way of thinking about questions of morality as he or she gets older.
The present discussion will provide a brief historical description of the theoretical foundation laid down by Piaget and Kohlberg; the method used by Kohlberg to assess moral development as well as alternative methods that have emerged more recently; some of the major criticisms and reconceptualizations of Kohlberg’s theory of moral development; and recent research that has pursued those criticisms in the areas of cultural differences, gender differences, and continued adult development.
THEORETICAL FOUNDATION
Kohlberg built on Piaget’s theory or cognitive development to hypothesize a sequence of six specific stages of moral judgment in individual development. This theory of moral development is based on a fundamental idea from Piaget that the way people think about the physical and social world is the result of an ”interactional” process between the human organism’s innate tendencies and influences from the environment.
This ”cognitive-developmental” approach is thus distinguished from both maturational and environmental theories of development. Maturational theories (Gesell 1956) maintain that patterns of behavior express the organism’s inherent tendencies. Development is seen as the natural unfolding of a process determined by hereditary factors. In contrast, environmental theories argue that behavior is determined primarily by external influences. From this point of view, behavior is not innately patterned but is essentially learned, whether as a result of conditioning processes that associate the behavior with particular stimuli, rewards, and punishment, or as a result of observing (and subsequently modeling) the behavior of others.
Social learning theory (Bandura 1977) has produced considerable research on how observational learning explains a variety of behaviors relevant to morality, including prosocial behavior (e.g., sharing, cooperation), aggression, resistance to temptation, and delayed gratification. More recent developments have pursued the question of how individuals exert control over their behavior, thus providing some balance to the theory’s focus on environmental influences. Bandura’s self-efficacy theory (1982), for example, emphasizes the individual’s expectations as important to the successful performance of a behavior. However, social learning theory has not addressed moral action and moral character in terms of a broad developmental course (Musser and Leone 1986).
Cognitive-developmental theory, on the other hand, focuses on the developmental process by which people come to understand and organize, or ”cognitively structure,” their experience. It attempts to resolve the ”nature-nurture” controversy by emphasizing the development of these cognitive structures as the result of the interaction between organismic tendencies and influences from the outside world. While particular ways of understanding experience may reflect innate tendencies, they develop in response to the individual’s specific experiences with the environment.
Thus, development is not seen as primarily maturational, because experience is necessary for cognitive structure to take shape. However, neither is development thought to be primarily determined by the environment. Rather, cognitive developmentalists argue that, because the underlying thought organization at each stage is qualitatively different, cognitive development is more than the progressively greater acquisition of information. Furthermore, at any given stage, the current cognitive structure can influence how the world is perceived. Thus, cognitive structure is seen to be ”the result of an interaction between certain organismic structuring tendencies and the structure of the outside world, rather than reflecting either one directly” (Kohlberg 1969, p. 352).
THE DOCTRINE OF COGNITIVE STAGES
Piaget’s theory of cognitive development maintains that cognitive structures are periodically transformed or restructured as they become unable to account for (or assimilate) new information from the external world adequately. These periods of restructuring result in new ways of understanding that are different from the earlier mental structures as well as from those to be developed later. This allows for the differentiation of distinct cognitive stages each identifiable by a characteristic approach to processing and organizing one’s experience of external reality.
Piaget (1960) identified four main characteristics of cognitive stages. Kohlberg (1969) maintains that these characteristics accurately describe his stages of moral development. The characteristics identified by Piaget are as follows:
1. Stages refer to distinct qualitative differences in the way a person thinks about an experience or solves a problem. Although the focus of attention may be the same, the mode of thinking about it is very different.
2. The progression of stages follows an invariant sequence in the development of individuals. That is, the order in which the stages occur is universal for all human beings. It is possible that the speed or timing at which one progresses through the stages may vary with individual or cultural environments—or even that development may stop at one point or another. However, a given stage cannot be followed by any other stage than the one that is next in the sequence. Conversely, the earlier stage must first be achieved before its inadequacies become apparent and the subsequent transformation to the next stage can occur.
3. The characteristic mode of thinking represents a structured whole. Specific cognitive responses to specific tasks depend upon the overall organizational framework within which one processes information. It is this underlying cognitive structure that produces a logical consistency to one’s responses. Thus, the stage is not identified by specific responses to specific stimuli, but it is the pattern in one’s responses that indicates a particular underlying cognitive structure.
4. The sequence of stages is hierarchical. At each stage, the underlying structure represents a more integrated and more complex organizational system, one that adequately accounts for information that had created discrepancies within the previous structure. For example, children in the preoperational stage of cognitive development (Piaget’s second stage) cannot understand that equal-sized balls of clay formed into two different shapes still have equal amounts of clay. However, children who have achieved concrete operational thinking (Piaget’s third stage) understand the principle of conservation and thus recognize that the amount of clay remains the same (is conserved) for both pieces, even though the pieces have changed in shape (Piaget and Inhelder 1969). The underlying cognitive structure of concrete operational thinking differentiates between amount and shape and integrates the information to achieve a more complex understanding of the phenomenon. It is thus logically superior to preoperational thinking. That the later stages in cognitive development are also more comprehensive and more advanced introduces a hierarchical element to the sequence. The stages of cognitive development are not just different but also hierarchical, in the sense that they provide a progressively more differentiated and more integrated—and hence more adaptive—understanding of one’s interaction with the environment.
KOHLBERG’S STAGES OF MORAL DEVELOPMENT
Kohlberg’s six stages of moral reasoning are divided into three levels, each consisting of two stages.
The three levels are differentiated according to what serves as the basis for the person’s moral judgment, specifically the significance given the prevailing, or ”conventional,” social expectations and authority. Briefly, the preconventional level, which is the level of most children under 9 years old, occurs prior to the individual’s achievement of a full understanding of what is expected or required socially. The conventional level, which characterizes most adolescents and adults, refers to an understanding of the social conventions and a belief in conforming to and maintaining the established social order. The post conventional level is reached only by a minority of adults, who understand and generally accept the social rules and expectations but recognize that these have been established for the larger purpose of serving universal moral principles. Should the social conventions conflict with these principles, then moral judgment at this level will support the principles at the expense of the conventions.
Within each level, the second stage is a more advanced form than the first. More specifically, the preconventional level refers to judgment based not so much on a sense of what is right and wrong as on the physical consequences that any given act will have for the self. Accordingly, at the first stage within this level, characterized by the punishment and obedience orientation, the child will make judgments on the basis of avoiding trouble. This includes obeying authorities to avoid punishment.
At Stage 2, still in the preconventional level, the individual has a sense of the needs of others but still makes judgments to serve her or his own practical interests. This is called the instrumental orientation. Although the person is beginning to understand that social interaction involves reciprocity and exchange among participants, moral judgment is still determined by the significance that the action has for oneself. Thus a child may share candy to get some ice cream.
Next, in the conventional level, moral judgment is determined by what is considered ”good” according to conventional standards. At this level, the individual has an understanding of what kind of behavior is expected. The first stage at this level (Stage 3) is characterized by the good boy-good girl orientation. Judgment as to what is right is based on living up to the expectations of others. It involves a trust in established authority and conformity for the sake of approval.
At Stage 4, the orientation is toward doing one’s duty. This is called the law-and-order orientation. The individual personally subscribes to the existing social order and thus believes that obeying authority and maintaining the social order are good values in their own right. Whereas behaving according to the social conventions is desirable at Stage 3 because it produces approval from others, at Stage 4 the individual has successfully ”internalized” these conventions, so that proper behavior is rewarding because it reinforces one’s sense of doing one’s duty and therefore produces self-approval.
At the postconventional level, one’s understanding of what is right and wrong is based on one’s personal values and a sense of shared rights and responsibilities. Morality is no longer determined simply by social definition, but rather by rational considerations. Stage 5 is characterized by the social contract orientation, which recognizes that conventions are determined by social consensus and serve a social function. There is an emphasis on utilitarian agreements about what will serve the most good for the most people. Here the person recognizes that rules or expectations are essentially arbitrary. The focus on agreement or contract produces an emphasis on what is legal and on operating ”within the system” to achieve one’s goals.
Stage 6, however, places the responsibility of a given moral decision firmly on the shoulders of the individual. The basis for moral judgment is found in universal ethical principles rather than socially established rules or expectations. One is guided by one’s own conscience and recognizes the logical superiority of principles such as respect for human dignity. At Stage 6, it is thus possible to adopt a position that is in conflict with the prevailing social order, and to maintain this position as morally correct.
In Kohlberg’s last theoretical paper, he and his colleagues attempt to articulate Stage 6 more completely (Kohlberg et al. 1990). They describe it as fundamentally characterized by a ”respect for persons,” specifically one that successfully integrates a sense of justice that is universal and impartial with an attitude of ”benevolence” that is em-pathic and understanding of the individual (see also Lapsley 1996).
MEASURING MORAL JUDGMENT
Kohlberg’s procedure for assessing moral judgment involves presenting a hypothetical ”dilemma” that requires the subject to make a moral choice. The most famous example refers to ”Heinz,” a man whose wife is dying of cancer. The woman could possibly be saved by a new drug, but the druggist who discovered it is charging an exorbitant amount of money for it, ten times what it costs him to make it. Heinz tried but could not raise enough money, so he steals the drug. Should he have done this?
Because Kohlberg’s scheme emphasizes cognitive structure, an individual’s stage of moral development is indicated not by the actual behavior that is advocated but rather by the pattern of reasoning behind the decision. Thus, two people may arrive at the same decision (e.g., that Heinz should steal the drug to save the life of his dying wife) but for two entirely different reasons. An individual at the preconventional Stage 2, operating within the instrumental orientation, might recommend stealing the drug because any jail term would be short and worth saving his wife. An individual at the postconventional Stage 6 might also recommend stealing the drug but with a different understanding of the dilemma: Although stealing would violate the law, it would uphold the higher principle of valuing human life and allow Heinz to maintain his self-respect.
The difference between the actual behavioral content of a decision and the cognitive structure of the decision is also illustrated when two people arrive at different decisions but for similar reasons. Thus, the decision not to steal the drug because Heinz would go to jail and probably not be released until after his wife died is also Stage 2 thinking. Even though the ultimate decision advocates the opposite behavior of what was indicated above, it is similarly based on the consideration of what would be most instrumental to Heinz’s own self-interest. On the other hand, an individual at Stage 6 might recommend not stealing the drug because, although other people would not blame Heinz if he stole it, he would nonetheless violate his own standard of honesty and lose his self-respect.
Because the stage of moral development is demonstrated not by the behavioral content but by the form of the moral judgment, the subject is allowed to respond freely to these moral dilemmas, and is asked to explain and justify his or her answer. The interviewer can probe with specific questions to elicit more information about the basis of the subject’s decision. Interviewers are trained to collect relevant information without directing the subject’s responses.
The subject’s answers are then transcribed and coded for stage of moral development. Kohlberg identified twenty-five aspects of moral judgment, basic moral concepts that refer to such matters as rules, conscience, one’s own welfare, the welfare of others, duty, punishment, reciprocity, and motives. Each of the twenty-five aspects was defined differently for each of the six stages of moral development. Originally, Kohlberg used an aspect-scoring system, whereby every statement made by the subject was coded for aspect and rated as to stage (”sentence scoring”). The subject’s usage of any given stage of moral reasoning was indicated by the percentage of his or her statements that was attributed to that stage. Aspect scoring also included an overall ”story rating,” whereby a single stage was assigned to the subject’s total response.
Coding difficulties led to the abandonment of the aspect-scoring system. Because the unit of analysis for sentence scoring was so small, coding often became dependent upon the specific content and choice of words and did not lend itself to identifying the general cognitive structure underlying the statement. Conversely, whereas story rating referred to the total response as the unit of analysis, it created some uncertainty when the subject’s answer included conflicting themes.
Kohlberg and his colleagues recognized these scoring difficulties and devoted considerable attention to developing a more reliable and valid scoring system. This led to ”standardized issue scoring,” which relies on the use of a standardized interview format. The subject is presented with three standard dilemmas, and the interviewer probes for only two issues that are specified for each dilemma (e.g., life and punishment in the Heinz dilemma). Scoring of the subject’s responses refers to a manual that describes the patterns of reasoning for Stages 1-5 on each issue (Colby et al. 1987). Stage 6 was dropped from the coding procedure, due to its empirically low incidence, but was retained as a theoretical construct (Kohlberg et al. 1990).
Because the focus of the new scoring system is directed more toward the abstract mode of reasoning, the unit of analysis is considered larger and less concrete than the single sentence. However, because this approach focuses on specifically identified issues, norms, and elements, it is considered more precise than the global story rating. Despite the qualitative nature of this approach and its potential vulnerability to rater bias, its developers report that long-term study of its inter-rater reliability, test-retest reliability, internal consistency, and validity has produced favorable results (Colby and Kohlberg 1987).
Validity has been a major concern regarding Kohlberg’s moral judgment interview. Kurtines and Grief (1974) criticized the low utility of moral judgment scores for predicting moral action. Other questions have been raised about the validity of the data collected, even for the purposes of assessing moral judgment. For one, use of the ”classical” dilemmas in this research has been criticized on grounds that they are not representative: Not only do they address hypothetical—as opposed to real-life—circumstances, but they refer to a limited domain of moral issues (e.g., property and punishment). Assessment may fail to indicate the extent to which the person’s moral judgment is influenced by the particular context provided by the dilemma. A related matter is whether responses are affected by the characteristics (e.g., the gender) of the story’s protagonist. Also the effect of differences in interviewing style, as interviewers interact with subjects and probe for further information, needs to be considered. Of particular importance is this method’s dependence on the subject’s verbal expression and articulation skills for the information that is collected. To the extent that the rating might be affected by either the amount of information that is provided or the manner in which it is expressed, the validity of the scoring system is called into question. (See Modgil and Modgil 1986 for discussion of these issues.)
An alternative to Kohlberg’s Moral Judgment Interview is the Defining Issues Test (DIT) (Rest 1986). This is a standardized questionnaire that presents a set of six moral dilemmas and, for each dilemma, specifically identifies twelve issues that could be considered in deciding upon a course of action. The subject’s task is to indicate, on a five-point scale, how important each issue is in deciding what ought to be done in the given situation. The subject also ranks the four most important issues.
Here, the term ”issue” is used differently than it is in Kohlberg’s new scoring procedure. The items are prototypical statements designed to represent considerations (e.g., ”whether a community’s laws are going to be upheld”) that are characteristic of specific stages of moral reasoning as they are described in Kohlberg’s theory. The importance assigned by the subject to items that represent a particular stage is taken to indicate the extent to which the subject’s moral judgment is characterized by that stage’s mode of thinking.
There are advantages and disadvantages to the DIT compared with the open-ended interview. Whereas the interview is helpful for originally identifying the considerations that may be relevant to resolving moral dilemmas, the DIT provides a more systematic assessment of the relative importance of such considerations. In the open-ended interview, it is never clear whether a specific concern is not identified because it is not important or because the subject failed to articulate it. Similarly, interviews are less comparable to the extent that subjects do not all address the same issues. These problems are avoided by the more structured DIT, because the task requires the subject only to recognize what is important rather than to identify and articulate it spontaneously. However, because recognition is an easier task than spontaneous production, it tends to allow higher-level responses. Another important difference is that the DIT measures the maturity of moral judgment as a continuous variable rather than in terms of the holistic step-by-step sequence of cognitive-developmental stages. Researchers must be aware of such differences when interpreting results.
A third instrument, the Moral Judgment Test (MJT) (Lind et al. 1981; Lind and Wakenhut 1985) similarly attempts to measure moral reasoning by the subject’s endorsement of specific items. Hypothetical moral dilemmas are presented, and subjects respond to a series of twelve statements for each dilemma. Each of Kohlberg’s six stages is represented by two statements, one in favor of and one against the particular action in question. Subjects indicate how acceptable they find each of the statements.
Citing Lind’s paper (1995) extensively, Rest and colleagues (1997) focus on an important distinction between the MJT and the DIT. Instead of adding ratings to indicate how much the subject prefers a particular stage’s statements (stage preference)— as most DIT studies do—Lind emphasizes how consistently the subject responds to different statements from the same stage (stage consistency). Lind argues that stage consistency is a more accurate measure of true cognitive structure, whereas stage preference is more indicative of an affective (like versus dislike) response.
Rest and colleagues (1997) use the DIT statements to construct a consistency measure that is similar to the one developed by Lind. They conclude that the stage-preference measure shows greater construct validity than the stage-consistency measures in differentiating groups with different expertise and different education. The stage-preference measure correlates more highly with moral comprehension—indicating longitudinal development—predicts both prosocial and antisocial behavior, and correlates with political attitudes.
The question of scoring for preference or consistency and what construct is measured by each approach is a legitimate methodological concern with important implications for our understanding of moral development. However, both the DIT and the MJT can be scored for preference and for consistency. Thus, they each remain a viable alternative for attempting to empirically measure moral judgment.
Another measurement tool is the Sociomoral Reflection Measure (SRM) (Gibbs and Widaman 1982; Gibbs et al. 1982), and a more recent variation is the Short Form (SRM-SF) (Gibbs et al. 1992; Basinger et al. 1995; Communian and Gielen 1995; Garmon et al. 1996). This is an open-ended, group-administrable instrument that asks subjects to rate the importance of such topics as keeping promises, affiliation, life, property, and law. It does not present specific dilemmas, but instead uses ”lead-in statements” that instruct subjects to generate their own example, such as ”Think about when you’ve made a promise to a friend,” prior to providing their rating of importance. The short form consists of eleven items that can produce a score ranging from 100 (exclusively Stage 1) to 400 (exclusively Stage 4).
Proponents argue that not only is the SRM-SF suitable for assessing stages of moral judgment, but because examples are self-generated, items also can be used to assess differences in content emphasis (Basinger et al. 1995; Garmon et al. 1996). As such, it is suggested as especially useful for research on cultural differences, gender differences, everyday life (versus hypothetical) experience, and the relationship between moral judgment and moral behavior (Communian and Gielen 1995).
As discussed below, perhaps the single most influential criticism of Kohlberg’s theory is Carol Gilligan’s contention that it fails to describe the moral development of females (1982). Her articulation of a more female-oriented ”morality of care,” complete with its own sequence of stages, has led to the development of the Ethics of Care Interview (ECI) (Skoe and Marcia 1991). Similar to Kohlberg’s methodology, the ECI assesses stage differences, but specifically as they are relevant to the development of care-based morality. Research with the ECI has recently been reviewed by Skoe (1998), demonstrating that the morality of care has important application to human development in general and to the development of personality in particular.
CRITICISMS AND FURTHER RESEARCH
Besides the methodological problems discussed above, Kohlberg’s theory of moral development has been criticized on a number of points. The major criticisms include the following:
1. The sequence of stages is more representative of Western culture and thus not universal or invariant across all cultures. Moreover, it is culturally biased in that it maintains the ideals of Western liberalism as the highest form of moral reasoning.
2. Like many theories of personality development, Kohlberg’s theory fails to describe the development of women accurately but provides a much better understanding of male development. This is a specific variation of the first criticism, suggesting that the theory itself reflects the sexism of Western culture.
3. Kohlberg’s theory fails to describe adult development adequately. In particular, its emphasis on abstract principles fails to recognize how adult moral judgment is more responsive to the specific practical matters of everyday, real-life contexts. Also, its emphasis on cognitive structure fails to recognize that changes in the content of moral reflection may be the most important aspect of adult moral development.
Cultural Bias. A cornerstone of cognitive-developmental theory is invariant sequence, the notion that the given developmental progression is universal for all human beings within all cultures. Because the conceptual organization of any given stage is considered logically necessary before the cognitive structure of the next stage can develop, each stage is said to have logical priority to subsequent stages. Shweder and LeVine (1975) take issue with both the notion of logical priority and the doctrine of invariant sequence, although they do not address the development of moral judgment per se. Specifically, they analyze dream concepts among children from the Hausa culture in Nigeria and conclude that there are multiple sequences by which such concepts develop.
Shweder (1982) follows up this initial skepticism with a fuller critique of what he sees as Kohlberg’s failure to recognize cultural conceptions of morality as relative to one another. He disagrees with the assertion that there is a rational basis upon which morality can be constructed objectively. Rather, he argues that the postconventional morality that Kohlberg maintains as rationally superior is simply an example of American ideology.
Similarly, others (Broughton 1986; Simpson 1974; Sullivan 1977) argue that Kohlberg’s theory is necessarily culture-bound, reflective of the Western society from which it originates. Simpson suggests that the specific moral dilemmas used in the testing situation may not have the same meaning for people of different cultures and thus the scoring system may not adequately detect legitimate cultural variations in moral structures. Thus, she maintains that the claims to universality are not valid. Sullivan goes even further, suggesting that Stage 6 reasoning is so rooted in the philosophical rationale for current Western society that it serves to defend the status quo. In doing so, it distracts attention from the injustices of such societies.
In an early response to the charge of cultural bias, Kohlberg and colleagues (1983) acknowledge the influence of Western liberal ideology on the theory. They agree there is a need to be more sensitive to cultural differences in the meaning attributed not only to the various elements of the research protocol but, consequently, also to the responses of the subjects themselves. However, they defend the claim to universality for the six-stage sequence of moral development and maintain that empirical research using the scientific method will help to determine to what extent this position is tenable.
They also maintain that, while it is appropriate to remain impartial in the study of moral judgment, this does not make it necessary to deny the relative value of certain moral positions. They assert that some positions are rationally superior to others. They thus continue to subscribe to the ideal that any given moral conflict can be brought to resolution through rational discourse.
Kohlberg’s position on invariant sequence has been supported by a number of cross-cultural studies, although postconventional reasoning (Stages 5 and 6) may occur less frequently in nonurbanized cultures (Snarey 1985). However, in a sample of subjects from India, Vasudev and Hummel (1987) not only found stage of moral development to be significantly related to age, but also found postconventional thinking to occur among a substantial proportion of adults. Concluding that commonalities exist across cultures, Vasudev and Hummel also suggest there is cultural diversity in the way moral principles are expressed, interpreted, and adapted to real life.
More recent research has increasingly acknowledged the significance of cultural influences on moral development. In another study with a sample from India, Moore concludes that ”the notion of justice, as defined by Western social scientists, was rarely used as a moral rationale” (1995, p. 286). Rather, results indicated moral judgment to be dependent on one’s religious ideology or social status. Similarly, Okonkwo’s study of Nigerians (1997) concludes that moral thinking and moral language are culture dependent. While some parallels to Kohlberg’s scheme were found, some ”well-articulated moral expressions could not be scored” (Okonkwo 1997, p. 117) due to the inability of Kohlberg’s instrument to adequately assess certain concepts that served as the basis for subjects’ moral judgments.
Ma and Cheung (1996), studying samples of Chinese, English, and American adolescents and young adults, likewise found cultural differences in the way subjects interpreted specific items on the Defining Issues Test. However, after they deleted some of the items used to indicate Kohlberg’s Stage 4, their samples demonstrated a consistent heirarchical structure across the three cultures. They thus conclude that, while different cultures may encourage different perceptions of specific moral statements, there is some support for the idea of a fundamentally universal development.
Markoulis and Valanides (1997) similarly addressed the cultural bias controversy in a conciliatory fashion. Comparing students from Greece and Nigeria, they found stage differences between the two cultures, but nonetheless found that the sequence of development was similar. Again, while cultural environment is recognized as a factor, invariant sequence in development is supported.
A related concern that has been receiving more attention from researchers is the question of whether differences in political ideology within a single, larger culture may be inaccurately represented as developmental variation. Specifically, Gross (1996) compared Americans who are pro-life on the abortion issue with others who are pro-choice, and also compared Israelis who disagree on now to handle the West Bank settlement issue. As long as socioeconomic status was similar, he found no difference between the relevant ideological groups and thus concludes that there is no evidence that ideological bias is built into the stages of moral development.
Conversely, Emler and colleagues (1998) argue that differences in moral development as assessed by the DIT more accurately reflect differences in political ideology. Thoma and colleagues (1999) acknowledge overlap between political thinking and moral judgment, but argue that Emler and colleagues and St. James (1998) provide no evidence to discount the DIT as a valid measure of moral development. Narvaez and colleagues (1999) attempt to resolve the issue with a model of moral judgment and cultural ideology as engaged in parallel development, each influencing the other to produce specific moral thinking.
Whereas recent times have been characterized by an increased sensitivity to cultural diversity and political ”correctness,” more attention has been drawn to the consideration of possible cultural and political bias in the theory of moral development. While some researchers have identified cultural differences in moral reasoning, this has led to an increased recognition of sociocultural factors in moral development (Eckensberger and Zimba 1997).
Shweder and colleagues (1987), for example, propose the social communication theory, which maintains that the learning of morality depends largely on the transmission of cultural ideology to children, by virtue of the evaluation and judgments that parents and others make. The point, of course, is that morality is socially constructed, not self-constructed (Emler 1998).
However, other researchers continue to maintain that, while we must make specific adjustments to our understanding of moral development, it is not necessary to abandon the general consideration of a single, universal pattern to human moral development. This area of inquiry thus promises to remain a controversial yet productive focus for several years.
Gender Bias. Carol Gilligan (1982) argues that the major theories of personality development describe males more accurately than females. She includes Kohlberg’s theory in this assessment and points to the prevalence of all-male samples in his early research as a partial explanation. Gilligan contrasts two moral orientations. The first is the morality of justice, which focuses on fairness, rights, and rules for the resolution of disputes. The second is the morality of care, which focuses on relationships, a sensitivity to the needs of others, and a responsibility for others. Gilligan asserts that the orientation toward morality as justice is especially characteristic of males and, conversely, that morality as care and responsibility is especially relevant to females. To the extent that Piaget, Freud, and Kohlberg each address morality as justice, they accurately represent male moral development but inadequately represent female moral development.
Gilligan argues that women are more likely to rely on the orientation of care to frame personal moral dilemmas. Furthermore, whereas the morality of care focuses on interpersonal relationship, it resembles the Stage 3 emphasis on satisfying the expectations of others. Gilligan believes this resemblance results in a high number of female responses being misrepresented with Stage 3 ratings.
Gilligan thus argues that Kohlberg’s theory and scoring system are biased to favor men. However, Walker (1984), after systematically reviewing empirical studies that used Kohlberg’s method, concludes that men do not score higher than women, when samples are controlled for education, socioeconomic status, and occupation. Similarly, Thoma (1986) reports that sex differences on the Defining Issues Test actually favor women but that the differences are trivial.
Kohlberg and colleagues (1983) address Gilligan’s criticisms and agree that the care orientation is not fully assessed by their measurement but disagree that this leads to a biased downscoring of females. They suggest that care and justice may develop together and that Stage 6 nonetheless represents a mature integration of the care and justice moralities (see also Vasudev 1988).
Walker and colleagues (1987) found that both the care and the justice orientations were used by both males and females. Furthermore, the orientation used was related to the type of dilemma being discussed. If the dilemma was focused on personal relationships, both men and women tended to use the care orientation. If the dilemma was impersonal, both men and women tended to express a justice orientation. This suggests that observed gender differences in moral judgment may be more a reflection of the particular kind of dilemma they choose to discuss. Perhaps females tend to report more relationship-oriented dilemmas, the kind that pull for care-basedjudgments (Yussen 1977).
Wark and Krebs (1996) show just this pattern: Females did not score lower than males on Kohlbergian dilemmas; however, females were more likely to report care-based dilemmas when asked to recall and describe moral conflicts from real life; this difference in the type of moral dilemmas accounted for differences in moral orientation from males and females.
Using the Sociomoral Reflection Measure-Short Form (SRM-SF), which does not rely on specific dilemmas provided by the researcher, Garmon and colleagues (1996) found support for gender differences in moral orientation, with females more likely to refer to a morality of care. However, they reject Gilligan’s claim of a bias against females. In fact, results (Basinger et al. 1995) have indicated a possible female advantage in early adolescence. As measured by the SRM-SF, moral judgment was found to be higher among young female adolescents than among their male counterparts. No gender difference was found in late adolescence or young adulthood. Communian and Gielen (1995) found similar results in an Italian sample, with early adolescent girls scoring higher than early adolescent boys, but no gender differences in adults. In another study of seventh and eighth graders, Perry and McIntire (1995) found that subjects used a care mode, a justice mode, and a third narrowly concerned ”selfish” mode to make moral decisions. The girls were more likely to use both the care and the justice modes, while the boys were more likely to choose the less developed selfish mode. Contrary to a bias against females, this research suggests that, at least in early adolescence, girls are more advanced in their moral development. Silberman and Snarey (1993) relate such a cognitive advantage to the earlier physical maturation of girls.
Consistent with the lack of evidence for a bias against females, Skoe (1995) found that Kohlberg’s justice-based moral reasoning was unrelated to sex-role orientation, as measured by the Bem Sex Role Inventory (Bem 1974). However, this research indicates an interesting pattern for care-based moral reasoning. Using the Ethic of Care Interview (ECI), Skoe found that care-based reasoning was higher in women who were more androgynous and who indicated higher levels of ego identity. Skoe concludes that women who relinquish the traditional female gender role are more likely to develop a mature care-based morality than are women who retain this role.
This would seem to be inconsistent with Gilligan’s argument (1982) that the morality of care depends on traditional female socialization. While this morality may be rooted in the traditional female role, Skoe’s findings suggest that its advanced development may require a more integrated, androgynous identity.
Current researchers seem to be recognizing that the different moralities go beyond simple gender role differences. Woods (1996) argues that both Kohlberg and Gilligan represent polarized, sexist views, limited to a focus on gender differences. She suggests that researchers need to take a more comprehensive view of the multiple biological and cultural variables that impact moral development, without reducing it to a discussion of sexism. Gilligan’s morality of care identifies a vital approach to morality that may have its origin and strength in feminine ideals, indeed that may be more salient in females than in males. However, women are not confined to it, nor is it confined to women, especially as gender roles become more relaxed.
Adult Development. A third major issue concerning Kohlberg’s theory is whether or not it accurately addresses continued adult development. This issue reflects a more general concern in lifespan developmental psychology regarding the inapplicability of Piaget’s model for cognitive development beyond adolescence, leading to a consideration of what has come to be called ”postformal” development (Commons et al. 1984). Murphy and Gilligan (1980) found that college and postcollege subjects not only indicated a greater tendency to appreciate the importance of specific contexts in real-life dilemmas but also indicated a slight tendency to regress from Stage 5 moral reasoning on the classical dilemmas. They suggest that a more mature recognition of the significance of contextual particulars leads one to question the validity of abstract moral principles (hence the regressed score). This argument is consistent with other work suggesting that adult cognitive development in general is marked by a greater appreciation of the practical realities of day-to-day living (Denney and Palmer 1981; Labouvie-Vief 1984; W. G. Perry, Jr. 1970). Related to this emphasis on the practical is the finding of Przygotzki and Mullet (1997) that elderly adults, when attributing blame, give more importance to the consequences of an action than to the intention of the perpetrator.
Finally, Gibbs (1979) argues that adult development is characterized more by increased reflection on such existential matters as meaning, identity, and commitment than by any structural change in the way the person thinks. Similarly, Nisan and Applebaum (1995) suggest that older adults give more weight to identity-related personal considerations when considering moral choices, unless they conflict with ”a moral demand” from an ”unambiguous law.” Gibbs (1979) suggests that Kohlberg’s postconventional stages are not structural advances over the earlier stages but would be more appropriately described in terms of existential development. In response, Kohlberg et al. (1983) maintain that Stage 5 represents a legitimate cognitive structure. However, they acknowledge the possibility of further nonstructural development in the adult years with regard to both specific contextual relativity and existential reflection. They suggest that such development could be described in terms of ”soft” stages that do not strictly satisfy Piaget’s formal criteria for cognitive stages.
SUMMARY
In spite of the formidable criticisms that have been levied against it, Kohlberg’s theory of moral development remains the centerpiece to which all other work in this area is addressed, whether as an elaboration or as a refutation. At the very least, Kohlberg has formulated a particular sequence of moral reasoning that adequately represents the prevalent sequence of development in traditional Western society. To that extent, it serves as a model, not only for building educational devices (see Modgil and Modgil 1986; Nucci 1989; Power et al. 1989), but also for comparing possible alternatives. Whether or not this sequence is in fact universal or relative to the particular culture—or a particular socialization process within the culture— is a debate that continues. Nonetheless, the scheme remains the prototype upon which further work in this area is likely to be based.
